32  Beappicker—On the Influence of Magnetism on the Rate of a Chronometer. 
VIII. 
[ArnoLp AND Dent. | 
The experiments made by Arnold and Dent (No. 17) have a special interest 
and importance, since they are the first attempt to separate the magnetic influence 
upon the balance from that upon the spiral spring. A further value lies in their 
being made by chronometer makers who had the advantage of a thorough know- 
ledge of the construction of a chronometer. This value would, however, have 
been considerably increaséd if Arnold and Dent had been acquainted with the 
previous experiments, especially those of Barlow and Harvey. 
They applied three bar-magnets, A, B, and C, each 16°12 ins. long, 1°68 in. broad, 
and 0:44 in. thick. They were lying in the meridian in the plane of the balance, 
and the chronometers alone were turned successively by 90°, or 180°, so that the 
figures XII, III, VI, and IX were exposed to the magnetic action. The distances 
were 7 and chiefly 3 inches; the time of exposure varied from 5 to 9 days. 
The temperature was read daily; but as the maxima and minima from the 
Royal Observatory Records are given, there is hardly any information to be 
got as to the changes of temperature to which the chronometers were actually 
subject. 
The time-pieces made use of were :— 
Chronometers Nos. 605 and 606, with gold spirals, and balances of silver, 
platinum, and brass. 
- Nos. 615 and 657, with steel springs and steel balances. 
< No. 600, with steel spring and platinum-silver balance. 
3 No. 274, with gold spring and steel balance. 
When the balance was at rest the cross-bar lay exactly in the line XII—VI; 
so that, as Arnold and Dent assume the poles of the balance to coincide with 
those of the eross-bar (which in the majority of instances seems to be the case), 
always either the north or the south pole of the balance must lie under the 
figure VI. Thus they state that the north end of the balance was lying under 
figure VI in the two chronometers Nos. 657 and 615. The same was apparently 
the case with No. 274. 
The results I reproduce in a condensed form, as I give only the means of 
Arnold and Dent’s figures. I remark that Nos. 605 and 615 were exposed 
simultaneously, the former to the north, the latter to the south pole of magnet 
